Keywords

18.1 Introduction

Shaped by a poor and uncertain natural supply the need to manage water resources on a collective basis has been a persistent political and social concern in the history of Spain. Responses to the increasing water gap have too often come from the supply side. Nevertheless, informal trades of water, mostly to meet short-run needs have also been pervasive (Hernández-Mora and De Stefano 2013). Formal water markets in Spain, however, are still incipient (Gómez et al. 2013).

Population and economic activity in the international Tagus River are concentrated in Madrid and Lisbon at both ends of the river basin district. Scarcity problems, though, are in principle more significant in the former.

Although the quality of surface water is relatively good in the upper basin, only one tenth of rainfall and runoff is available to cope with more than two thirds of the demand for urban uses in the entire watershed. In clear contrast to that, the intermediate reach of the Tagus River basin is dominated by the use of water for extensive agriculture (TRBA 2014), disputed by the more productiveFootnote 1 irrigated agriculture in the Segura River basin (SE Spain), close to the Mediterranean coastline and which is connected to the Tagus by a major diversion project (Tagus-Segura Water Transfer).

Within the last three decades water management in Madrid has been a clear example of a gradual adaptation towards a more efficient use of infrastructures, together with pricing schemes and other incentives designed to adjust water demand (TRBA 2014). However, in the two decades before the current downturn, the combination of intense demographic change, economic expansion, and rapid and extensive urban development pushed to the limit the capacity to manage an increasing water demand within the range of available resources and current water infrastructures.

Since the early 1990s, voluntary agreements to transfer water use rights from agriculture to urban uses emerged as an alternative to cope with the recurrent water supply deficit during a number of dry periods since then (especially the two drought events of 1990–1991 to 1994–1995, and 2004–2005 to 2007). This option gained social and political momentum since the costs of the best available alternatives already in place (efficiency improvements, use of strategic reserves, additional water works) grew in the margin (Estevan and Lacalle 2007). In fact, efficiency in water treatment and distribution in Madrid is already above 80 %; a high percentage of wastewater is currently being re-used for watering public gardens, for high-pressure street washing and to maintain environmental flows. In addition, some strategic groundwater reserves have only been used (when not strictly preserved) for extreme events, but their water stock is limited by definition. Finally, the construction of new infrastructures in the Tagus has been ruled out because of its economic and political cost (TRBA 2014).

This is the context where pioneer voluntary agreements, albeit actively supported by the water authority, to transfer water between irrigation districts and water utilities and sprung up for the first time in Spain. Needless to say that, as above, in a drought-prone and water-scarce country as Spain, informal water transfers have always spontaneously occurred and even some meaningful trades have taken place amongst farmers. The difference in the latter derives from the volume of exchanges, the parts involved in the bargaining process, the purpose of exchanges, and their importance to foster the adaptation of the institutional framework in order to allow for a wider use of water use trades.

This chapter aims at illustrating the performance of a diversity of water transfer arrangements in which economic incentives were used to tackle challenges posed by drought events. Two different transfers are to be assessed: the public water utility (Canal de Isabel II, CYII) taking water (for which it holds rights) from the Alberche River to supply Madrid city’s domestic uses (also including the water transfer from Las Parras stream in the Middle Tagus to the Alberche Canal to compensate irrigators); and the water right transfer from the Henares Canal irrigators to the Mancomunidad de Aguas del Sorbe (hereafter MAS) (Sorbe River Water Community) to supply water and sanitation to different towns in the Henares’ Corridor. Each of them is full of nuances and different features, which will shed light on formal and informal practices of water exchanges.

18.2 Setting the Scene: Challenges, Opportunities and EPIs

The study site is located in the upper section of the Tagus watershed and includes parts of the Alberche, Jarama and Guadarrama catchments (three of the main tributaries of the Tagus), an area of 8,022 km2 (see Fig. 18.1). It includes the densely populated Madrid metropolitan area (4,567,190 inhabitants) (Eurostat 2014) and its sprawl along the Henares Urban Corridor, and the two irrigated areas involved in the assessed water trades, respectively located in the Alberche catchment and on the banks of the Henares River.

Fig. 18.1
figure 1

The study site (Source: Own elaboration)

Full size image

In the pre-crisis period 1996–2008 Madrid increased its real GDP by more than 50 %, its number of employed people by more than 1.25 million and its real GDP per capita from EUR 19,755 to EUR 23,636, thus generating a pull effect which increased population even during the first year of the current economic crisis (INE 2014). The municipalities of the MAS, with most of its served population and industry located in the Region of Madrid, showed a very similar trend, with high economic and population growth in the towns and villages of the Henares Urban Corridor (mainly in Alcalá de Henares and Guadalajara). These trends propelled water demand and gave rise to the two water trades assessed.

Regarding the first trade, the Alberche Canal irrigation area spreads out in 10,000 ha, with an approximate irrigation demand of 75 million cubic meters per year (hm3/yr). Irrigation water supply depends on the Picadas Dam, connected to the Valmayor Dam, operated by CYII. Due to the joint effect of scarce water inputs and consumption rises in the preceding years, water supply for Madrid city could have actually become uncertain.

In March of the hydrological year 1993–1994 water storage in the reservoirs of the Alberche System was less than 128 hm3, almost equal to the volume of the Alberche River flow allocated for Madrid’s supplyFootnote 2 (119.8 hm3). Given the hierarchy of use in the Spanish Water Act (i.e. the priority of domestic supply) and the legal water allocation of the CYII in the Alberche, this implied that irrigated crops depending on the water of the Alberche Canal would not have received enough water for the cropping season (TRBA 2008). CYII used their entitlement to supply Madrid city; to compensate local farmers, a water transfer was conceived from the middle stretch of the Tagus River (Las Parras Stream) to the Alberche, exclusively for irrigation purposes.

As to the second trade, the MAS was created to supply water to the towns of Alcalá de Henares, Guadalajara and other municipalities of the Henares River valley, downstream Beleña’s reservoir. Total population supplied was 68,000 inhabitants (CYII 2011), and the volume of water used was 6.8 hm3 (MAS 2014). The project for the aggregation of water and sanitation services in the area included growth projections that have been largely exceeded. These included a maximum population for Alcalá de Henares and Guadalajara of 100,000 inhabitants each and 25,000 for the other municipalities at stake for the purposes of this assessment. Nowadays, total population supplied by MAS is 363 126 inhabitants (INE 2014) (which includes the seven original municipalities and other six that joined later) plus 20,000 inhabitants from municipalities which are not MAS members. In terms of supply for water and sanitation services, that can be translated into an increase of the total water use, which in 2005 was already at 46 hm3 (MAS 2014).

MAS held water rights for 1,300 l/s and water supply mainly stemmed from the Beleña Dam, with only 50.3 hm3 of effective capacity (Estevan and Lacalle 2007). By the end of the 1990s it was already supplying water to a much larger population than originally projected (TRBA 2014). The river had an average contribution of 168.68 hm3/year (TRBA 2008). The Sorbe River system demands an average of 75 hm3 per annum from the Beleña Dam, out of which 51 hm3 are for domestic supply, leakages amount to 12 hm3 on primary and secondary mains, 9 hm3 help maintain the environmental flow and the remaining water went to filtration and evapotranspiration (CYII 2011).

Before 2001–2002, MAS managed to provide services to all municipalities every year. However during that hydrological year, Sorbe’s contributions were 13 % of average levels during the twentieth century. At the beginning of February 2002 the reservoir level was under 9 hm3, equivalent to 2 months of consumption. Given the risks for spring and summer seasons, and at the request of the TRBA, MAS contacted the irrigators of the Henares Canal to negotiate the purchase of a certain amount of water, within the framework of article 67 of the Water Law (on lease contracts).

The Henares Canal has a total irrigation area of 7,500 ha placed in 15 municipalities at southeastern Guadalajara. It holds water rights up to 5,600 l/s from the Henares River and an upper bound of 66.18 hm3/year. The dams of Palmaces (with a total maximum storage capacity of 31 hm3), Alcorlo (180 hm3), and El Atance (35 hm3) regulate the canal. An additional advantage was that the canal is parallel to MAS pipes and is just 2 km away from the Mohernando wastewater treatment plant (WWTP).

18.3 Water Markets in Action

Nowadays both transfers are deemed successful examples of drought adaptation. They are part of the institutional developments that have moderately boosted the use of voluntary transfers of water use rights as a water security mechanism avoiding other costly or politically challenging alternatives such as new water infrastructures or new expensive water sources.

However, sharp increases in water demand revealed the need for a more flexible approach (the Spanish model is based upon a concessional regime, not a water market at such), so that users could meet their demand but not necessarily at the expense of a higher use of the resource. This was even more compelling considering that water administrative mechanisms to re-allocate water (such as administrative procedures for water rights expiry, water concession audits, etc.) proved to be ineffective. For instance, the river basin authority would take a year and a half to process water right applications, even when water was available. Lease contracts, on the other hand, allow water users to get exclusive water rights in just 2 months (Vázquez 2010).

18.3.1 Main Outcomes of the EPI Implementation

18.3.1.1 Environmental and Economic Outcomes

High demographic and economic growth have led to increased water demand and the upturn of physical capital in the upper and middle stretches of the rivers of the region: as a result of that, Madrid is now amongst the Spanish regions with more heavily modified rivers (Alcolea and García 2006) and, in turn, in the world (Gómez 2009). No significant surface water supply increases can be obtained through additional human-made capital, and in fact there have not been new investments on relevant dams since the 1970s (CYII 2011). While demand increased and surface resources grew unable to meet water demand during droughts, groundwater has been used as a buffer stock. Although some strategic reserves have been kept to couple with drought events, overall groundwater withdrawals during these years have scaled up, exhausting aquifers and leading to a poor ecological status (TRBA 2014). Under these circumstances, the solution to scarcity problems came through significant rises in water efficiency that made it feasible to bridge the gap between water use and withdrawals.

Within this scheme, economic policy instruments such as water rights transfers were called to play a key role provided they did not contribute to exacerbate increasing demand trends. In the first water trade analysed, water from the Middle Tagus compensates irrigators in the Alberche Canal, which in turn provide water to guarantee Madrid reserving water flows, crucial for the provision of water and sanitation services to the households in the Madrid Region. However, the lack of a formal previous agreement with the irrigators of the Alberche Canal to replace water from the Alberche River with water from the Middle Tagus sub-basin hampered the right operation of Las Parras-Alberche Canal connection, as it happened in the 2004–2005 drought, thus leading to overexploitation of Alberche River’s resources (TRBA 2008). Moreover, while the Alberche-CYII water transfer supplied high quality freshwater to Madrid, irrigators from the Alberche complained about the low quality of the water diverted from Las Parras stream.

Unlike in Madrid city, in the MAS water supply problems were related to a lack of regulation capacity as compared to increasing water demand. There had been flawed attempts to upgrade the regulation capacity of the system. Several proposals had actually been analysed to transfer water to the Sorbe catchment from the Alcorlo Dam (180 hm3) in the Bornova catchment (TRBA 2008). However, the procedure was slow and the last investment alternative (a transfer of 80.9 hm3/year from the Beleña Dam to the Alcorlo Dam) is still today at a standstill due to its high environmental impact. Consequently, the solution had to come from a water right transfer from the Henares Canal (TRBA 2014), which provides water of a lower quality than that contained for example in the Alcorlo Dam. To some extent, this is an idle and derelict infrastructure, initially planned to transfer water to the Beleña Dam in several undone projects.

In turn, the Henares Canal takes its resources from the Henares River. Economic and population growth in the surrounding areas had increased water demand in the Henares catchment, thus leading to a reduction of water supply guarantee during scarcity junctures. As a result, groundwater abstraction in the area had increased, and although the quantitative status of the Guadalajara aquifer was (and is) still fair, its qualitative status was poor (TRBA 2014). Indeed, non-point pollution resulting from agricultural activities is still a problem of major concern, also affecting the Alcorlo Dam; its resources are used for urban supply as well as the provision of amenities (TRBA 2014).

In both cases, the water transfer directly responds to the need to guarantee water supply to the population. Indirectly, it also contributes to guarantee water supply to other urban uses, such as the service sector, the most relevant economic sector in the study area (INE 2014). Water trading in both areas has been only made possible through the development of water infrastructures to transfer water from agricultural districts to urban areas.

In the first water trade assessed, the Alberche – CYII water transfer has been working since 1967, although the expansion that is relevant to our analysis was installed in 1993. The expansion allowed the CYII to use 119.8 hm3/year at a cost of 10 billion pesetas (equivalent to 67 million ECU1993 Footnote 3). The Region of Madrid was responsible for the payment, which had to be effective within a period of 25 years. The compensatory water transfer from Las Parras stream in the Middle Tagus sub-basin to the Alberche River was built in 1991 with an initial capacity of 5 m3/s. It was used for the first time in 1993 when, following a drought, 35 hm3 were transferred from the Middle Tagus to the Alberche Canal. In 2006, the TRBA allocated an additional amount of 100 hm3 of the Alberche’s resources to the CYII water utility.Footnote 4 This increase was followed by an expansion in the capacity of Las Parras – Canal de Alberche water transfer, up to 7 m3/s at a cost of EUR 2 million. In 2008, the Las Parras-Alberche infrastructure had to be used again to solve a water shortage in the Alberche River. The cost of the intervention was EUR 1.48 million, and the water transferred had a lower quality than that of the Alberche River. An additional projected measure consists in the modernization of the irrigation systems in the Alberche Canal, with an estimated cost of EUR 50 million, which is expected to save 25 hm3/year through more efficient irrigation.

In the second water trade (MAS), the limited capacity of the Beleña Dam fostered an agreement between the irrigators of the Henares Canal and the MAS to transfer water rights for urban water supply. The agreement entered into force in February 2002 and transferred the use of 20 hm3/year. from the irrigators of the Henares Canal to the MAS. This agreement was extendable for 2-year periods up to a maximum of 10 years. The infrastructural cost amounted to EUR 3.5 million and mainly consisted in the construction of the Maluque-Mohernando connection, with a length of 2 km and a capacity of 1.3 m3/s. The maximum flow rate was variable throughout the year: between September and April the transfer could work at its maximum capacity, but from May to August the maximum flow was 300 m3/s. The fixed costs of the water transfer for the MAS was EUR 38,000/year, plus EUR 0.01 for the first 4 hm3 and EUR 0.02 from that amount onwards. During the summer months (June–August) each additional hm3 was paid at EUR 0.03/hm3. Besides, MAS had to pay the pumping costs to the TRBA, which during 2005–2006 amounted EUR 388,000 (CYII 2011).

The guarantee of water provision through the use of EPIs such as water right transfers has contributed to consolidate economic growth in the most dynamic urban area of Spain: Madrid and its sprawl along the Henares Urban Corridor. Coupled with economic growth, urban water productivity in the Madrid Region has experienced a remarkable growth during the period 1997–2006. The service sector (80.5 % of region’s GDP) more than doubled its water productivity in this period. Overall, water productivity has increased in the secondary and tertiary sectors as GVA and GDP grew (INE 2014), showing a consistent pattern that can be described as Verdoorn’s LawFootnote 5 for water (Pérez-Blanco and Thaler 2014). In those sectors, the apparent productivity of water is well over EUR 1,000/m3, while in the building sector it shoots up to over EUR 13,000/m3. On the other hand, irrigation productivity is under EUR 1/m3 for many crops in the Talavera and La Campiña agricultural districts (roughly corresponding to the areas supplied by the Alberche and Henares canals, respectively), which show an average water productivity of EUR 3.57/m3 in La Campiña and EUR 3.79/m3 in the Talavera agricultural district (INE 2014; MAGRAMA 2009).

Although transfers from agricultural to urban sectors unambiguously result into higher water productivity values, these results need to be taken with caution. Agriculture may be a minor economic sector for the aggregate of the study area but in certain rural areas it is actually the main activity, and reducing water availability could lead to substantial economic losses and depopulation in these areas. Noteworthy, rainfed agriculture shows much lower income than irrigated agriculture in the two agricultural districts at stake (MAGRAMA 2009).

18.3.1.2 Distributional Effects and Social Equity

Farmers growing irrigated crops in La Campiña Agricultural District get an average income of EUR 1,123.06/ha, with important variations across crops. Corn, which covers the widest area, produces an average income of EUR 2,000/ha while barley obtains less than 900 and peas less than 200. These three crops altogether represent 95 % of the study site area; the remainder of the area is covered by more profitable and water demanding vegetables. As an indicator of the value of water, it can be said that the average income amounts to EUR 0.19/m3, but 23 % of the irrigated area using 30 % of the water might be generating an income lower than EUR 0.04/m3 (MAGRAMA 2009).

The importance of water for irrigation can be observed if comparing the previous numbers with those of rainfed agriculture, covering around 120,000 ha and earning EUR 490/ha on average. For example, as per the most common crop, both under irrigated and rainfed agriculture, irrigation facilities and water represent a shift from EUR 460 to 2,000/ha of income and an increase from 2,300 to 11,000 kg of average yield (MAGRAMA 2009). From a social viewpoint, water does make a difference in the study site indeed.

The average income obtained in the Talavera agricultural district (Alberche catchment) is about twice that of the Henares (EUR 2,180/ha) and although cereals still account for three quarters of the irrigated area, crops are more diversified than in the Henares. Average income is EUR 0.32/m3 with lower variations than in the Henares. The dominant crop is also corn, which covers 40 % of the irrigated area, uses 45 % of water, and yields EUR 0.22/m3 (MAGRAMA 2009).

The assessed water use right trades have not had significant impacts on material living standards, since water was guaranteed for both water supplying irrigation districts holding stakes. From stakeholder consultation, it can be inferred that the irrigators of the Henares Canal did not suffer noteworthy losses (Gómez et al. 2011). The fact that farmers in the Henares Valley accepted to give their water up at a price lower than EUR 1/m3 is but an indication that probably (at least part of) those water resources were not being used at all for irrigation.

As above, irrigators from the Alberche Canal complained about the low quality of the Middle Tagus water received via Las Parras stream. A study carried out by the public utility (Estevan and Lacalle 2007) states that the conductivity of the Tagus River up to Talavera can reach 2,000 μS/cm, which basically means that it is semi-brackish water. This does not seem to have led, though, to critical production losses or major protests.

This idea that no major equity impacts were found is reinforced by the fact that compensation payments were implemented in both water transfers (in the Henares – Sorbe transfer, as part of contractual explicit clauses). In 1993, the Region of Madrid paid for the energy costs of pumping (50 million pesetas; that is, 0.335 million ECU1993) (Estevan and Lacalle 2007) in which the irrigators would incur to divert water (35 hm3) from the Tagus River to the Lower Alberche Canal trough Las Parras stream. In the agreement the irrigators did contract the energy supply, which was thus partially paid by the CYII. Despite a number of interviews with stakeholders (representatives from the River Basin Authority, on one side, and CYII water public utility, on the other), no significant evidence has been obtained as to why this compensation was implemented in spite of the utility holding water rights in the Alberche. It seems part of a compromise between the company and the basin authority.

Regarding the transfer from the Alberche to supply Madrid city, households did not face an increase in water tariffs due to additional expenses for the public utility (power for pumping).

In November 1996, the energy company Unión FENOSA claimed the payment of 1 billion pesetas (6.7 million ECU1996) from CYII arguing that some production losses would occur after the construction of the Alberche – CYII transfer. Since the transfer started working, Unión FENOSA considered that its water concessional rights were being affected. The transfer reduced the volume of water that the company could turbine, and hence, its capacity to generate electricity. These damages were estimated at roughly 1.2 billion pesetas (7.3 million ECU1996). CYII managers argued that they were acting under the safeguard of the water rights they held, granted by the TRBA, in which there was neither specific constraints to the transfer nor compensations to other third parties.

In terms of positive equity impacts, local communities served by MAS and benefitting from the water transfers from the Henares River, managed to elude water restrictions. There are records of some complaints regarding the quality of water, but this was always in accordance with regulations. MAS, on the other hand, does not only supply households but also industries; for some of them water is an essential input.

18.3.2 The EPI Setting Up

18.3.2.1 Institutions

A significant number of institutions were involved in the implementation of the two analysed water transfers. The first transfer (that from the Henares Canal to the MAS) describes a situation in which the transfer itself was formally feasible. The other (CYII using their entitlements to supply Madrid with water from Alberche River) is an example of a situation where the water transfer is viable but only under more specific circumstances. The former is a case in which a water right transfer is performed in strict sense. The latter, on the contrary, is an example of a water right holder (the public water utility) using their rights and the affected irrigators being compensated through a decision by the TRBA.

In the first transfer, the irrigators of the Alberche CanalFootnote 6 do not hold rights although they are beneficiaries of the allocation of public water flows for irrigation. These farmers have not been granted a formal entitlement and thus their rights are not registered, which implies that they cannot be part of a water transfer contract. Nevertheless, as an exception, the RDL 15/2005 allowed water users adjoin to public irrigation land to sign lease contracts provided some conditions were met (BOE 2005). Resources allocated for this irrigated area (hydrological plan, 1999) were 7,500 m3/ha and year.

In the transfer from the Henares to the Sorbe, it is interesting to point out that although in those two events in which water right transfers were needed they were implemented following the conditions required by the MAS, the contract did not include a single binding clause for the Henares Canal irrigation community to set aside water resources for water transfers; neither did it contain any provision that could shed light on the relative priority of water leasebacks or trades over the Canal users’ risks (Estevan and Lacalle 2007).

18.3.2.2 Transaction Costs and Design

It should be clear that water transfers generate benefits and costs relative to what the situation would have been in the absence of these transfers (in other words, as compared against a baseline scenario). If stakeholders had not been aware of those impacts and there had been no compensation for those bearing negative impacts, there would have been a clear economic incentive to politically contest these water transfers. Should that have been the case, the effort that stakeholders would have devoted to lobby decision-making should have been added to the explicit costs of the EPI implementation; these costs are generally assumed away. Yet, there was compensation in both cases, as it has been explained in previous sections. Compensations followed a number of meetings which entail non-negligible transaction costs.

For the first trade, in March 11th, 1993, the Bo of the Irrigation District of Alberche River held a meeting to analyse alternative solutions to provide water for irrigators of the Lower Alberche Canal. Because of the urgency in building the Alberche – CYII connection, the transfer started working in November of that very year.

As per the second trade (2002), the agreement was signed in February 8th, 2002 and in July of the same year it started working for 4 months. The connection between the Henares Canal and Mohernando’s treatment plant did not work again until June 2005 for a period of 8 months (up to January 2006). The fact that it took less time to implement the second transfer is an indicator of lower transaction costs, due to the prior effort, which provided some institutional assets for ulterior water trades.

There is no evidence of the time devoted to these meetings, since, as above, these two water transfers were designed as ad-hoc urgent measures to tackle drought consequences. However, despite this emergency feature, the process was longer (and consequently transaction costs higher) than one could infer. Once irrigation was established in the Alberche, different drought periods threatened water availability for farmers. This motivated, in 1991, the construction of an emergency infrastructure, thanks to an intake from the Tagus. This means that although no significant transaction costs may be linked to the decision to transfer water in 1993, some ex-ante costs may need to be taken into account (no available information has been made available to the authors), regarding the construction of the infrastructure for water transfers.

18.3.2.3 Implementation

After centuries of enforcement of the appurtenance principleFootnote 7 a number of factors led to the transfer of water rights among users. Economic development and urban growth, as well as droughts themselves, could no longer be managed within the limits imposed by a water law based on the needs of a formerly agricultural society and the limits of traditional administrative procedures. As a result of these parallel developments Spanish water legislation was upgraded (1999) to allow for the transfer of water rights.

The Spanish scheme accepts the transfer of water rights, but this faces somewhat significant restrictions. One of them is that trades may only take place between and among uses of similar or superior ranking in the hierarchy of uses.Footnote 8 The ranking of preferences in the law is based on custom (BOE 2001), thus hampering efficiency gains that may be attained through exchange, since it does not take account of water productivity in different alternative uses.

The drawbacks of the system are best illustrated by the fact that legislation issued to cope with droughts, and related emergency measures include exceptions to the above-mentioned order of preferences. If allowances are needed for the system to perform there is an indication that structural, permanent rules, may need some amendment if water markets are to be developed.

Whilst the coordination between parties (the public water utility and the irrigators from the Alberche Canal on one side, and the MAS and the irrigators from the Henares on the other), was realistic, especially for the facilitating role of the Tagus River Basin Authority, some flaws can be observed regarding the implementability of these transfers in other contexts. Under current water legislation and institutional set-up, request for a transfer may be approved by default, if the administration does not approve or refuse it within 1 or 2 months, depending on the reach of the transfer (within or outside the same community of users). That legal provision provides a powerful incentive for administrators to reject transfers outright should they be complex and time consuming.

Water trading, as assessed in these two water trades, may face additional challenges. Allowing transfers from agricultural to urban uses (as in the two cases that have been assessed) may bring to the negotiation process water resources that are not being effectively used, unless strict monitoring provisions are implemented. Not surprisingly, given the low quality of soil in the region of Madrid, agriculture is a waning activity and in some areas water allowances are higher than the effective demand for irrigation water. Once subsidies from the Common Agricultural Policy (CAP) have been phased out and agricultural markets have been liberalized, the irrigation sector in some areas may be in excess water supply.

Both trades were the result of measures to cope with droughts. Because of this, all procedures and public works were executed on an emergency basis (time laps between decisions and effects were 8 months for the 1993 transfer, and 3 months for the 2002 one), which hampered whatever participation process. This lack of participation, though, does not seem to have damaged the acceptance of water transfers, nor has it conditioned its design. Yet, these instances have created a favourable context (built on protocols and formal procedures), which opens up space for further transfers.

Despite the fact that no participation took place regarding the assessed transfers, there have been ulterior consultation processes within the context of the water planning drafting stages. As part of those, a number of potential conflicts were identified (regarding the ecological status of river Tagus, as pacing the town of Talavera; the availability of water resources in the Alberche – mainly water for irrigation; the use of Sorbe’s resources; etc.).

18.4 Conclusions

The emergence of water markets, as many other once innovative EPIs, is a gradual adaptive and learning-by-doing process that must be judged by its ability to push water institutional development rather than by the failure or success of the experience itself. The water transfer from the Alberche River although useful to manage the supply deficit in the 1990s would not be an alternative nowadays anymore and many doubts exist as to the real prospect of repeating the 2002 water trade from the Henares Canal to the MAS in the same formal terms. The actual value of these examples is in the lessons that can be drawn and its importance to furthering agreements on reallocating water use rights as an instrument for water security.

Both examples also illustrate the critical importance of managing water use conflicts. It is well known in economic analysis that water management is essentially conflict management. In fact, according to the Spanish law, households have a priority over irrigators in water use, and there is no need for a voluntary agreement to take water away from farms in order to guarantee a sufficient supply of drinking water in dry periods. The real buffer for drinking water in Spain is the irrigated agriculture whose use rights are defined every year depending on the rainy season. Moreover, instead of just taking water or forcing farmers to let water flow, the agreement is easier to reach if alternative resources are available, the harvest is protected and third-party effects are avoided.

This is the real meaning of the 2002 transfer. The existence of these alternative resources is precisely what makes the replication of this trade almost impossible in 2011 (as there is evidence of overallocation or water rights in the middle Tagus river). Nevertheless, lessons learnt can be important to understand how, instead of paying for water, agreements are easier to reach when alternative sources are provided to guarantee existing uses, particularly in irrigated agriculture. Nowadays, alternative resources can either come from re-used or desalinated water.

Water trading also faces some important challenges. Allowing transfers from agricultural to urban uses may bring to the negotiating table water resources that are not being effectively used. In fact, given the low quality of soil in the Madrid area, agriculture is a receding activity and in some areas water allowances are higher than the effective demand for irrigation water. Paradoxically, once EU subsidies for agriculture have been phased out and agricultural markets have been liberalized, the irrigation sector in some areas may be in excess of water supply. The fact that farmers in the Henares valley accepted to give their water up at a price lower than one eurocent per cubic meter is but an indication that probably those water resources were not being used for crops. Hence, water trading might not be a means to reduce water scarcity but rather to increase it and would not be instrumental to re-allocate water but to effectively increase its use. This would be a real risk should water saved after the publicly supported shift towards more efficient irrigation systems, becomes part of the water trading system rather than being left in already degraded aquifers.

Both transfers were designed, as emergency measures, for severe drought situations that threatened water supply of important cities and towns including Madrid. It is therefore evident that a more systemic consideration of non-structural alternatives to water management in Spain, rather than a drought-based-emergency resort to market-like solutions, may be requested on economic efficiency grounds.